Crescent Fracture (SI Joint Fracture-Dislocation, Day Classification)

What it is. A crescent fracture is a fracture-dislocation of the sacroiliac (SI) joint and a recognised subset of lateral-compression pelvic-ring injuries (Young-Burgess LC, typically LC-II). The mechanisms are high-energy: road-traffic collisions, falls from height and crush injuries.

How often. Calafi and Routt identified 129 crescents in 128 patients over 7 years at a busy pelvic referral centre, with a mean age of 42 years and a mean Injury Severity Score of 25.5. The injury is uncommon, but a regular attendee at pelvic trauma units.

The company it keeps. Associated injuries are the rule. In the original operative series (Borrelli 1996, 22 patients, mean age 25), 64% had other associated injuries and 23% a closed head injury, yet all were haemodynamically stable at presentation despite the rotationally unstable hemipelvis.

Both sides. Bilateral crescent fractures are rare but reported (Bachhal 2021 case-based review), so always check the contralateral side on the CT.

The anterior variant. The anterior sacroiliac fracture-dislocation (ASFD, the "locked pelvis") is easily misdiagnosed as a lateral-compression crescent. In one Korean series it accounted for 25% (15 of 59) of crescent fractures, and it needs anterior plate fixation rather than iliosacral screws.

Mechanism. A lateral compression force applied to the pelvis disrupts the SI joint posteriorly, and the injury propagates as a fracture through the posterior iliac wing. That fracture splits off a crescent-shaped fragment of ilium which remains attached to the sacrum by the strong, intact posterior sacroiliac ligaments. The crescent is the posterior component of a Young-Burgess lateral-compression injury and is usually accompanied by an anterior ring (pubic rami) injury.

Stability. Because the posterior ligaments stay attached to the crescent fragment, the injury is generally rotationally unstable but vertically stable. The word is generally: in Zong's series 4 of 31 (12.9%) iliac crescent fracture-dislocations were vertically unstable.

The L5 root. The L5 nerve root runs over the sacral ala roughly a finger-breadth (around 2 cm) medial to the SI joint, where it joins the lumbosacral trunk between the ala and the L5 transverse process. It is vulnerable to the original fracture displacement and, intra-operatively, to an anterior SI-joint plate, a reduction clamp or a misdirected screw, so identify and protect it in any anterior approach. The full plexus anatomy is in the Lumbosacral Plexus Anatomy topic.

Day grades the injury by the size of the crescent fragment and the proportion of the SI joint that is dislocated. The smaller the crescent, the more of the joint is dislocated.

View larger

Why the type matters. The classification is useful because it directs the approach and the reduction technique. The larger the crescent, the more the posterior stability resides in the iliac fragment, so fixing the iliac fracture restores the SI joint. A small crescent leaves little ilium attached, the injury behaves like a near-complete SI dislocation, and iliosacral screw fixation across the joint is needed. Type II sits in between.

Its limits. Some crescents do not fit neatly into the three types, and the scheme has been proposed for expansion.

Radiographs. An AP pelvis with inlet and outlet views. Look for:

• the posterior iliac fracture line
• SI-joint diastasis or step-off
• vertical migration
• the associated anterior ring (pubic rami) injury

CT is essential. It defines the size of the crescent fragment and the proportion of the SI joint dislocated, which together give the Day type. It also shows any sacral or anterior-ring injury and the sacral osseous corridor for screw planning. For the anterior variant, CT is essential to diagnosis and preoperative planning (Trikha).

Telling the anterior variant apart. Chun (2024) found that ASFD shows greater vertical displacement on the outlet view (mean 9.5 vs 1.9 mm) and a larger pelvic asymmetry ratio on the AP view (mean 7.8 vs 4.1) than the LC-II crescent. In 6 of 15, iliac override blocked SI reduction.

The whole patient. This is a high-energy pelvic-ring injury. Screen for:

• haemodynamics, because a vertically unstable subset bleeds
• associated visceral and urogenital, head, chest and abdominal injuries
• neurology of the lumbosacral plexus and roots
• the skin, for a Morel-Lavallée lesion over the flank or trochanter

When stability is in doubt. Examination under anaesthesia or stress views if rotational stability is unclear after static imaging.

Resuscitation first. Follow ATLS and pelvic-trauma principles: a pelvic binder or sheet for the rotationally unstable ring, haemorrhage control, and a search for the associated injuries that dominate early mortality.

Who bleeds. Most crescents are haemodynamically stable; all 22 in the Borrelli series were. The vertically unstable minority in Zong's series were all haemodynamically unstable on admission, with more associated injuries and higher Injury Severity Scores, so instability on imaging changes the resuscitation posture.

Temporising. When definitive surgery must wait, external fixation holds the ring, and transcondylar skeletal traction can hold the SI joint reduced until delayed ORIF is safe (Zong). Time definitive fixation for when the patient is physiologically ready and the soft tissues permit.

The soft-tissue envelope. An overlying Morel-Lavallée lesion, a closed degloving over the flank or trochanter with a real infection risk, influences both the timing and the posterior approach to fixation. It is managed as in the Morel-Lavallée Lesion topic.

The goal. The principal goal of surgery is accurate and stable reduction of the SI joint, restoring the posterior ring to allow weight-bearing and to prevent malunion and chronic pain. The Day type then selects the approach, and whether to fix the fracture (the ilium), the joint (iliosacral screws), or both.

Definitive fixation by type. The Day type decides the construct:

• Type I (large crescent): open reduction and internal fixation of the iliac fracture through an anterior or posterior iliac approach, with plates and/or lag screws, which re-establishes SI-joint stability. If closed reduction is achieved, percutaneous screws perpendicular to the fracture line are an alternative (Shui).
• Type II (intermediate): reduce the joint and stabilise both the fracture and the SI joint as needed, often iliosacral screws with or without iliac fixation, or crossed percutaneous fixation (Shui).
• Type III (small crescent): treat it as an SI dislocation, with closed (or open) reduction and then percutaneous iliosacral screw fixation across the SI joint.
• Anterior variant (ASFD): anterior SI plating.

The shift to percutaneous fixation. Many crescent fractures are amenable to closed reduction and percutaneous iliosacral screw fixation. Calafi and Routt fixed 60% of crescents with percutaneous iliosacral screws after closed or open reduction, and Shui's comparative series supports percutaneous-first when closed reduction is achievable.

Other options. In selected stable patterns, non-operative management or external fixation has a role. The choice is individualised to the fracture type and the patient.

Posterior ORIF of the ilium. The approach is the posterior extrapelvic approach over the iliac wing (Borrelli). The patient is positioned prone or lateral. Both the fractured iliac wing and the SI-joint dislocation are reduced anatomically, the crescent fragment being levered back onto the sacrum. Pointed reduction clamps are used, and a laminar spreader in the joint allows it to be inspected and cleaned.

Borrelli's construct. Intertable lag screws, placed between the inner and outer tables of the ilium and perpendicular to the fracture, plus an outer-table neutralisation plate.

Iliosacral screws: reduce first. A percutaneous iliosacral screw fixes a reduced SI joint; it does not achieve the reduction. It cannot be placed safely through a malreduced joint, because a malreduction throws the bony corridor out of line and pushes the screw towards neural structures. Obtain an accurate closed (or open) reduction first, then fix.

Stay in the corridor. The screw runs from the lateral ilium, across the SI joint, into the S1 (and sometimes S2) vertebral body, and must remain entirely within bone. A transiliac-transsacral screw, passing through both ilia and across S1, needs a corridor wide enough to accept it. An extraosseous screw injures the structures just outside the corridor:

• the L5 nerve root, crossing the ala superiorly
• the S1 foramen and roots
• the sacral canal

The dysmorphic sacrum. A substantial minority have upper-sacral dysmorphism, which narrows and angulates the S1 corridor so that a transsacral S1 screw may be impossible. Use an oblique S1 and/or an S2 corridor instead. Recognise it on the preoperative outlet view and CT:

• an acute alar slope
• mammillary bodies
• a residual S1 disc
• a "tongue-in-groove" SI joint
• a raised colliculus

Proving the position. True inlet and outlet fluoroscopic views plus a strict lateral sacral view, which shows the iliac cortical density and the alar slope, or navigation, confirm that the screw stays in bone. Type III crescents are the classic indication, and Shui used iliosacral screws for all Type III injuries.

Anterior plating for the ASFD variant. Here the iliac fragment dislocates anterior to the sacrum and may override it, blocking closed reduction. The approach is ilioinguinal or modified Stoppa, with a standard anterior SI-joint plate. The L5 root crosses the ala directly in the plating field, so identify and protect it.

Early complications are bleeding, nerve injury and wound trouble:

• Haemorrhage and instability, mainly in the vertically unstable subset
• Iatrogenic L5 root injury from an anterior plate, reduction clamp or misdirected screw over the sacral ala
• Wound problems, and the infection risk of an overlying Morel-Lavallée lesion

Infection. Borrelli reported one late iliac-crest osteomyelitis, at 6 months. Shui found infection rates significantly lower with percutaneous than with open fixation.

Malreduction. A malreduced SI joint leads to chronic posterior pelvic pain, SI nonunion and post-traumatic SI arthrosis, and to pelvic malunion with limb-length and rotational deformity. Malreduction, not the bone loss, drives the long-term morbidity, which is why the goal is anatomical reduction.

Other late problems. Fixation can fail, and implants can irritate:

• Loss of reduction, if fixation understates the instability (the missed vertically unstable pattern)
• Hardware irritation: prominent screw heads or plates may need removal

The classification in practice. Day's 16 patients managed according to the classification achieved good functional results at approximately 2 years. The authors' own caveat is that the SF-36 and MFA summary scores were confounded by the associated high-energy injuries.

ORIF. All 22 of Borrelli's patients healed clinically and radiographically within 8-10 weeks, with no acute wound, neurologic or vascular complications.

Percutaneous fixation. In Shui's 117-patient comparison, the closed and percutaneous group had less blood loss, shorter surgery and hospital stay, and better functional scores. There was no difference in reduction quality, revision rate or iatrogenic nerve injury.

The vertically unstable subset. It does well when recognised: all 4 of Zong's patients rated good or excellent on the Kobbe system after posterior ORIF.

The anterior variant. All 59 crescents in Chun's comparative series, both groups, achieved union after type-appropriate surgery, with no residual vertical-displacement difference once reduced and fixed appropriately.

The bottom line. Accurate SI-joint reduction is the determinant the surgeon controls. The associated injuries are the main confounder of reported functional outcome, and residual disability is driven largely by them rather than by the crescent itself.

• No crescent-specific guideline exists. The injury is managed under general pelvic-ring injury principles (AO/OTA 61-C / Young-Burgess LC-II) within ATLS-style resuscitation frameworks; the Day classification is the operative planning tool, not a guideline.
• The global practice shift is visible in the literature: routine ORIF (Borrelli 1996) → percutaneous iliosacral fixation where closed reduction is achievable (Calafi & Routt 2013, 60%; Shui 2015, 73 of 117) → recognition of variants (anterior ASFD needing anterior plating; vertically unstable patterns needing ORIF-level stability).
• Classification extensions are being proposed internationally: a Chinese 4-type regimen for SI dislocation (Tan 2017) incorporates Day's three crescent types as its Type III subtypes - evidence that Day's framework is the global reference point even where schemes are expanded.
• Registry reality: crescents sit inside pelvic-ring registries rather than having dedicated databases; reported outcomes are dominated by small single-centre retrospective series, so the evidence base on this page is the literature, not a registry.